#1

I'm a mad fan of pluck as many may know - but the samples can be a little iffy.
Having spent a bit of time now using tidalcycles and estuary (extra limitations on samples), I've got some ideas about how I'd like a sample set to be - ie

My theory (so far) around what makes a good set for an analog harmonic instrument sample set is:

  1. Pitch the sample at Concert C
  2. Make sample:0 a "commonly" used technique of the instrument
  3. Extra samples should maintain the pitch, but vary the technique of playing the note based on commonly used technique variations of that instrument

I've developed my first sample set for fretless electric bass using this methodology, hopefully addressing some of the issues found in pluck - I'd love to get some feedback/critique so I can hone it a bit before I move onto more sets, so please give it a try!

11 Likes
#2
2 Likes
#3

These are some wonderful samples! The variations of playing techniques make a much more dynamic sound. It also sounds super nice through reverb.

Here's a video of how I used them

1 Like
#4

That sounds absolutely killer! what a showcase - thanks for giving them a try :pray:

2 Likes
#5

Thanks. I was searching a way to combine the different type of samples with different notes

(edit1)

d1 $ somecyclesBy 0.1 (hurry 2) $ fast 2 $ shuffle (wchoose [(1,1), (4,0.051)]) 
    $ note (scale "mixolydian" "0 <5 [~ 2]> <4 1> <1 10>" - 5) 
	# s (wchoose [("flbass:1", 2), ("flbass:5", 0.51), ("flbass:6", 0.1), ("flbass:7", 0.1), ("flbass:8", 0.1)])
	# shape 0.1
	# legato (range 0.4 0.6 rand) 
	# gain 1
	#octave 5

d2 $ fast 2 
	$ every 4 ("q" <~)  
  $ s "{kick:11 <snare:2 snare:2>,  hc hc clap:2*2?0.5}"
  # gain 1 
  # shape 0.2

d3 $ fast 2 $ s "kick:13 snare:4"
# gain 1.1

d4 $ loopAt 1 $ slice  8 "0 [1 4] 2 3 <4 2*2> 5 6 7" $ s "break:4"
# gain 1.3

d5 $ somecyclesBy 0.2 ("s" <~) $ somecyclesBy 0.2 ((arp "updown") . (fast 2)) $   n ("<c'min7 bes'dom7>" + "<0! -5!>") # s "superpiano"
# gain 1

Just let us know when the slap bass with the pops and slaps are ready :wink:

2 Likes
#6

Thanks to everyone who's been trying this out! With so many sample sets out there I'm stoked that I've managed to make something that people are interested in

I made a major update this morning - I've added some detail to my formula:

Problem: I've supplied 1 octave only (excepting the harmonics), as these get pitched they lose/gain length which makes it a bit unpredictable/inconsistent in terms of sound

To try and solve this, I've done two things:

  1. Add another octave set (c3)
  2. Make the long/short note length consistent. I picked 0.25 cps as a reference point (60bpm @4/4) - long notes are 2 cycles in length (8s) which allows for a natural decay/release, and short notes are a quarter cycle in length (1s)

This will chop and change as cps' change of course, but at least there is a consistent reference point you can calculate back to.

Apologies, you'll need to change your patterns to adjust for the shifted samples - which was the other thing I did:

The samples are currently grouped by octave:
flbass:0-5 = C2 length and attack variations
flbass:6-11 = C3 length and attack variations
flbass:12-14 = harmonic/ghost notes

I haven't found a reason to do this grouping by style/octave/other yet, so I just went with this - I'd love to hear any reasons people think it should be one way or the other - sooner rather than later if possible, I'm conscious that every change can bugger up peoples workflows :wink:

That's the next cab off the rank - I have a 5 string jazz bass waiting for the call up... however I'm far from an experienced slap player, so it's going to take some practice :wink:

1 Like
#7

Had a play this morning... I'm not really liking the short samples anymore, I think I'll revert/change them - has anyone else tried them at 1s length?

Update: I've gone back to a much shorter sample for the short set - I'll leave it from here I think.
Thanks again to everyone who gave them a try :slight_smile:

Update:

famous last words... I've just added some palm mute samples... NOW I'm done :stuck_out_tongue:

1 Like
#8

nice work, these sound great :slight_smile:

https://www.instagram.com/p/CJwFkGoBR9A/

1 Like
#9

I've also had a go with these fantastic fretless samples. Thank you so much! This experiment kinda just wrote itself, owing to the inspiring bass sounds... Please enjoy!

https://www.instagram.com/p/CJtP1WLDawb/

setcps 0.53

d1 $ slow 2 $ s "samples-flbass*20" # n (run 20) # speed "<[1 0.75] [1 1.2]>" # hpf 180

d7 $ cat [
  "{bd}%5"
  ,"bd*2 ~ bd ~ ~"
  ,"bd*2 ~@4"
  ] # n 0 # speed 1 # coarse 9 # gain 1.15 # pan 0.45

do
  let kick = every 4 (degradeBy 0.9)
        $ every 4 (stut 3 0.4 (-1/20))
        $ n "<0(7,10) 2(6,10)>" # s "super808" # gain 0.9
        # sustain (slow 2.2 $ sine)
        # voice (choose [0,1,1.5])
  let hat = chew 10 "0 1 2 3? 4 5 6 7 8 9*4"
        $ stut 2 0.6 (-3/20)
        $ degradeBy (segment 10 $ slow 2 $ saw)
        $ s "<0*20 10*10>" # s "superhat"
        # gain 0.85 # sustain "{0.2@4 1 0.1}%20" # cut 1 # pan 0.3 # room 0.1
  let snare = every 3 (rev)
        $ every 4 (stut 3 0.2 (-1/20))
        $ cat [
          s"{~ <1 1*2>}%5"
          ,s "~"
        ] # s "supersnare" # gain 1.15 # room 0.15 # decay (choose [0.2, 5]) # pan 0.6
  d2 $ (|* gain 1.1)
    $ stut "[8 1@9]/5" 0.6 (-1/500)
    $ whenmod 10 8 (juxBy 0.6 ((3/10) ~>)
      . (# cutoff (rangex 1200 16000 $ fast 5.5 $ sine)) . (# lpq 0.1)
      . stut 3 0.1 (3/20)
      . slow (choose [2,4]))
    $ stack [
        kick
        ,hat
        ,snare
      ]

d5 $ (# pan (segment 1 $ slow 1.01 $ sine))
  $ slow 2 $ cat [
    n "c'min"
    ,n "f4'maj"
    ,n "ef4'maj"
    ,n "af4'maj"
  ] # s "supersquare"
  # gain 0.4
  # room 0.7
  # voice 0.3
2 Likes
#10

Thanks for checking them out, that's a beautiful sounding pattern and an unexpected (for me) use of the samples, I think will need to explore further - thanks for sharing! :grin:

#11

7 posts were split to a new topic: What if there was a sample pack format with metadata for SuperDirt?

#13

@yaxu first question after reading through the thread and looking at your samples as reference:

Automatically topped and tailed and normalised to -3db with ecasound/sox

I've normalised mine to -1.0 db... is this something we should be making consistent from the outset?

#18

Some flbass samples:

1 Like